PCMark 10 is getting battery life and cross-platform Office benchmarks

If you go by sales data, most folks on using PCs are on laptops, yet there's really no standardized laptop battery testing methodology. Sure, we have our own testing style, but it differs from, say, Anandtech's, and their differs from Tom's Hardware's…and so on. Underwriter's Laboratories (UL), the group that purchased Futuremark back in 2014, just announced that PCMark 10 is getting two new benchmarks—one of which is a battery life test.

UL says that the PCMark 10 battery life benchmark will be a realistic simulation of practical, day-to-day usage patterns. Specifically, it will test four types of battery life: office work, video playback, gaming, and idle time. 

The benefits of having a standardized battery life benchmark could be pretty large, allowing readers of publications that use the benchmark (or individual users) to compare battery life of various laptops across websites. Of course, that all depends on whether the benchmark is in fact an accurate gauge of the true longevity of the machine; we all have choice things to say about how representative (or not) 3DMark can be.

The other benchmark on its way to PCMark 10 is the "Applications" benchmark. This test will run on both x86-64 and ARM machines and will directly gauge performance in Word, Excel, PowerPoint, and the Edge browser, making it a less synthetic test than PCMark10's standard "Productivity" benchmark. The company says that results will be comparable across systems and architectures.

UL says the PCMark 10 battery life test will hit at the end of March, while the Applications benchmark will come by the close of the first half of 2019.

Comments closed
    • e1jones
    • 10 months ago

    Tried to run it on the backup computer… it failed every time on the first part (video conferencing). LOL 🙁

    Oh… and Steam… yuck.

    • Stochastic
    • 10 months ago

    While it’s nice to have a standardized battery life test, comparisons between publications will only be possible if things like screen brightness, Wi-Fi signal strength, etc. are controlled. I hope this makes reviewers’ jobs a little easier, though!

      • ikjadoon
      • 10 months ago

      Yes. Screen brightness validation is a must. Screens consume upwards of 30% of a laptop’s battery energy in light to medium use.

      I’m really curious how they’ll standardize that, if at all. WiFi, too.

      This would be a great time for laptop manufacturers to offer brightness tables (20% brightness = 125 nits). Maybe PCMark can force the benchmark to only run within a certain nit range for validated scores.

      Glad to see idle time listed there. Damn tired of the hot bag issue.

      • DavidC1
      • 10 months ago

      Screen brightness can be tested on the reviewers side. They will tell you “120 nits” or something like that.

      Wi-Fi signal strength is too much of a minefield. Every laptop will have different Wi-Fi strength and standardizing it means suddenly its no longer realistic. Perhaps it should be applied elsewhere too? Test it at same CPU performance, memory bandwidth, storage performance?

      It’s not like PCMark is pioneering or anything though. We already have MobileMark. The 2014 version seems to be on the optimistic side when comparing it to reviews. They used to be quite realistic back when MobileMark 2012 was the norm.

      Not to mention the general benchmark suite for PCMark sucks. I basically skip the PCMark section when looking at reviews.

        • ikjadoon
        • 10 months ago

        But not many reviewers will measure the nit level, which makes it harder to compare as not all reviewers get the same configuration (some with 1080p and an i5; others only the 4K panel and an i7).

        One or two nit brightness ranges would solve a lot of problems.

        WiFi levels should be standardized so reviewers can be more consistent.

        I don’t understand your point. Standards mean a set environment; obviously, one environment isn’t 100% realistic for every user, but it should capture a wide range of users.

        Your other points are unfortunately missing the point of benchmarks. They benchmark the OEM’s hardware and software…not the environment.

        MobileMark 2014 sets weak guidelines. “Please promise not to run this below 150 nits. But set it to whatever you want above that.”

Pin It on Pinterest

Share This